How the rise of open access is altering journal publishing

The scholarly journal has taken handily to the web over the last decade, in a way that scholarly books have only begun to explore tentatively. This transformation of journal publishing is giving rise to new and various economic forms, with perhaps the most radical and promising innovation taking the form of open access. Open access is a term which has been in use since 2002 to describe those research articles that readers can find online and read in full without being a member of a subscribing library or otherwise having to pay a fee.¹ At this point, the number of papers published in either open access journals or posted in open access archives (with more on these two models below) is growing, and, in 2009, somewhat more than 20 per cent of the peer-reviewed literature published that year, no matter where it was published, could be found in open access formats online (Björk et al., 2010). The number of peer-reviewed, open access journals indexed in the Directory of Open Access Journals is nearly 10,000 titles, with many more still to be listed there (Edgar and Willinsky, 2010).

Yet open access is not just a child of the digital age but the latest expression of long-standing principles of scholarly publishing having to do with the openness of science (David, 1998). These principles hold that the value and quality of research and scholarship are related to the extent of its circulation, as greater dissemination subjects the knowledge to greater review, as well as enabling more people to take advantage of it. Which is to say that open access represents a considerable step towards the scholarly ideal of universal access to the knowledge needed to advance knowledge, and one that follows on the move of most journals to online editions – a move which has already advanced this principle of increased access and circulation.

In terms of scientists’ reading habits, for example, Tenopir and King have recorded a general increase in the number of articles read in an ongoing study dating back to 1977, with a significant leap taking place since the beginning of the twenty-first century (2008).² Yet within this set of online articles that encompasses the better part of the research literature, those articles that have been made open access, whether through archiving or by the journal, appear to be read more often in the short term (Davis et al., 2008) and cited more frequently with time (Hitchcock, 2012), suggesting that they have indeed increased their contribution or value within the literature by virtue of being open access. Additionally, reader interest in open access articles has been on a rapid increase since 2003 (Laakso et al., 2011).

As for apprehensions that work made freely available will not be valued, the open access journal PLOS Biology set out to allay some of those concerns and, within two years of its 2003 launch, managed to score biology’s highest Journal Impact Factor (which is based on the number of citations per article). Journal Citation Reports consistently ranks PLOS Biology top of the biology category as a highly-cited journal in this field (Björk, 2012). PLOS Biology is part of the Public Library of Science (PLOS), which publishes a total of seven open access journals. These journals charge an article processing fee, which authors pay out of their research grants (although waivers are granted based on need). In the biomedical field, in which PLOS publishes, paying such fees is legitimate and feasible given existing grant structures.

In a more recent development, PLOS has led the way in developing the new ‘mega journal’ with PLOS ONE, which began in 2006. PLOS, financially self-sustaining since 2010, releases 14,000 articles annually through PLOS ONE, with the emphasis on the rapid publication of ‘good enough’ research and post-publication evaluation through article-level statistics. Papers submitted to PLOS ONE go through a peer-review process that concentrates on objective and technical concerns to determine whether the research has been sufficiently conceived and executed to justify inclusion in the scientific record. Unlike many journals, PLOS ONE does not use the peer-review process to determine whether or not an article reaches a particular level of ‘importance’ as that is left to the readers to decide, with article-level metrics, from citation to blog posting, reflecting readership response (PLOS, 2012). Articles from PLOS ONE now make up 1.5 per cent of the articles in PubMed (ibid.). Since the launch of PLOS ONE, other publishers, notably those with commercial interests, have followed suit with their own mega journals.³

Open access has also had an impact on some of the best-established titles such as New England Journal of Medicine, which now offers its content to readers without charge six months after publication (Kennison, 2005). Open access can be said to offer advantages for new journals, increasing their value, especially for faculty members who are building their reputations. How much more quickly does a journal become known when its content, from the very first issue, is made immediately available to interested readers everywhere, with Google Scholar indexing it, with citation counts for each article (the evidence cited in the previous paragraph points to greater readership and citation levels for open access research).

Open access is also increasing the value of knowledge through a number of other channels, including publishers’ generous agreements that provide developing countries with free or deeply discounted access to a wide array of journals (Willinsky, 2006: Chapter 6), the use of open access research by government policy makers (Willinsky, 2003) and by professionals in other fields (Willinsky and Quint-Rapoport, 2007), and the contribution that open access is making to the educational quality of public knowledge resources such as Wikipedia (Willinsky, 2008).

Further still, the late-2000s recession led to a reduction in funds for library collections, as university incomes across the country declined (Henderson and Bosch, 2010).⁴ Increasingly, university libraries are struggling to maintain their commercial subscriptions and are either questioning or discontinuing many of them. As a result, a ‘downstream effect on the scholarly society and commercial publishers who rely upon institutional subscriptions and licenses for revenue’ has surfaced. This has made for ‘significant and positive changes in the prospects for changing funding streams to support open-access publishing’ (Eckman and Weil, 2010).

In this chapter, we pursue the value of open access as a matter of the economics by which this research and scholarship are being made freely available across a highly stratified journal market. This market – which has been estimated to run to US$8 billion annually for English-language research in science, technology and medicine alone (Ware, 2009) – can be said to be made up of three relatively distinct publishing economies, comprised of independent journals, scholarly society publishers and commercial publishers.⁵ All three segments publish peer-reviewed journals of varying quality in virtually all of the disciplines. What sets one segment apart from another is the cost structure historically associated with its publishing activities. In an odd economic turn, then, the price charged for a journal is more closely associated with the segment that publishes it, whether independent, scholarly society or commercial, than with either the quantity or quality of the journal’s content. Using the per-page costs for journals, based on the subscription fees paid by libraries, Bergstrom and Bergstrom have found that commercial publishers charge three-to-nine times as much as society publishers across six disciplines from ecology to physics (2004: 897). As for quality, they used the price-per-citation, finding that the citation, as a measure of reputation, could cost ten times more with commercial publishers (ibid.).⁶ And it is this legacy of stratification that shapes the economics of open access today. In light of that, let me say a little more about each segment at this point.

At this point, all three segments are looking at open access in various forms, by adopting author self-archiving policies, by utilizing open source software systems, by charging article processing fees and by providing delayed open access, with more on these below. The number of fully open access journals is growing at approximately 15 per cent each year, as new journals are founded and subscription journals move to the open access model. By contrast, subscription journals are growing at around 3.5 per cent (Laakso et al., 2011).

While the motives may differ, there is a shared sense that the web is giving rise to growing access expectations, especially around public goods such as federally sponsored knowledge. As the International Association of STM Publishers, which represents 110 societies and commercial publishers in 21 countries, has put it: ‘Publishers are currently either implementing, or testing, ways to harmonize this visionary goal [of open access] with economic reality’ (International Association of Scientific, Technical and Medical Publishers, 2008: 10). The ‘economic reality’ alluded to here is marked by stratified journal pricing practices among scholarly societies and commercial publishers, reviewed above, and further complicated by thousands of independent journals that charge neither reader nor author.¹¹ When publisher associations speak of harmonizing this visionary goal with economic reality, or when they speak of their willingness to support ‘all business models – all models that ensure continuity and sustainability of the journal model that have brought such significant insight and information to the scientific community’, the message is clear (Mabe, 2008). If the world wants open access, then open access it will be, on the publishers’ terms. That is, the economics of open access will have to be capable of sustaining current revenue levels and price-increase rates. Two things are troubling about this picture, despite the wonderfully promising aspect of universal access to research and scholarship. The publishers want to sustain a stratified economy that (a) has yet to be justified in any substantive way, and (b) comes with a total price tag that the academic community has declared on more than one occasion to be ‘unsustainable’ (Byrne, 2003; Davies and Greenwood, 2004; Leonard et al., 2004). If the academic community cannot keep up with the price of open access, then we are again faced with what we have now: limits to the circulation of knowledge.

The economic discrepancies within this three-tiered unsustainable structure may arise out of recent history, but they speak to the lack of governing economic and intellectual property principles that might guide the handling of this public good, principles that address the mix of academic and business interests, the balance of public and private concerns, and the need for open and closed account books in scholarly publishing. This is not the place to consider these wider principles. However, its analysis of the current economics of open access is intended to have the value-add, we might say, of bringing these economic discrepancies to the fore and placing pressure on this need for principled distinctions to be made in how this public good is shared and distributed.

The independent origins of open access

Open access emerged in the 1990s as tinkering researchers discovered that they had everything to gain by using networked computers to share freely their work and that of others. They posted papers on their websites, set up a place to share papers about to be published, and established what we now think of as open access journals. By taking advantage of these new technologies, they proved to be sources of innovation and development in scholarly publishing. They could: (a) demonstrate the value of having immediate and unrestricted access to recent papers and work in progress; (b) exercise rights of academic freedom and editorial independence in setting up new journals; and (c) bring journals published by scholars working outside of the industrialized world into far wider circulation, as well as bringing such scholars into editorial positions with journals that could be edited from anywhere.

Let us look now at two early and important examples in demonstrating the original impulses and economic principles. In 1991, the physicist Paul Ginsparg began a pre-print service in his field that is now known as arXiv.org at Los Alamos National Laboratory, which virtually initiated the author self-archiving model of open access that publishers in all three segments support to a degree. Two years later, in 1993, the statistician Gene Glass launched Education Policy Analysis Archives, a peer-reviewed journal at Arizona State University, providing a model of independent open access journal publishing.

Ginsparg’s pre-print server – to which physicists submit their working and published papers, as well as stop in on a daily basis to consult new work – continues to provide a means of opening up what had long been the private circulation of pre-prints and off-prints to all physicists, establishing, in effect, what has evolved into the author self-archiving route to open access. He simply took hold of the networked computer and used it to vastly extend the reach of a pre-print culture intended to improve the circulation of knowledge, with arXiv.org now offering up to half a million papers in physics, mathematics, computer science, biology and finance.

Similarly, Glass was able to see that email together with the networked computer greatly facilitated the two key requirements of scholarly publishing, namely managing the peer review of papers and then publishing the resulting work by circulating it electronically.¹² The open access journal ran on this new ability to bring the entire publishing process within the scope of what was sitting on and under the scholar’s desk.

At the outset, at least, Glass and Ginsparg did not have scholarly societies to sustain nor corporate shareholders to appease, nor did they have overhead expenses or staff payrolls to meet. They had only an interest in and an idea about furthering the scholarly circulation of knowledge. They did not imagine that ‘information just wants to be free’, but they understood that the resources they needed to make this knowledge available fell within the scope of the work sponsored by government, foundations and (largely tax-exempt) institutions in the name of learning.

Their innovative approaches to open access were, however, based on their own particular technical skills. In a way that had obviously not been the case with printing, the tools needed to manage, distribute and circulate research were now part of the scholarly trade. While others have continued to build their own archiving and journal publishing systems, what made these two forms of open access economically viable on a global scale was a new and related development in how software was produced. Out of a similar spirit of what was known, programmers were coming together to produce free and open source software, beginning in the 1980s and leading to such prominent instances as Linux, Apache and Mozilla Firefox (Weber, 2004). Around the turn of the millennium, special open source systems were developed specifically for managing eprint archives (e.g., EPrints, DSpace, Fedora/Fez) and for publishing journals (e.g., DPubS, HyperJournal, Open Journal Systems).

Self-archiving open access

After having set up three discreet segments within the scholarly publishing market, we now need to complicate things, as different approaches to open access cut across these divisions, without fundamentally altering, I will argue, the economic stratification. Independent journals, scholarly society publishers and commercial publishers have, in large measure, accepted an author’s right to self-archive a copy of the work that has been published in a journal. In the longest-standing instance of arXiv.org, for example, the publishers of the relevant physics journals report that after more than a decade of a substantial portion of their content being archived they have experienced no greater decline in journal subscriptions than they are experiencing more generally, as the total number of subscriptions sold by individual journals declines as a result of the proliferation of titles and the increases in prices (intended to compensate for subscription losses).¹³ Self-archiving was also the first of the open access initiatives to benefit from the development of open source software systems, in the form of EPrints (http://www.eprints.org/) – released by the University of Southampton in 2000, and which made freely available the means for libraries and others to set up such archives (Tansley and Harnad, 2000).

Out of recognition that at least some authors were archiving their work on their own websites, and that this seemed unlikely to hurt subscriptions, publishers began to include a right to do so within the copyright transfer policies that authors were asked to sign. At this point in time, 65 per cent of the publishers (among the 1063 that have responded) grant authors the right to post a peer-reviewed version, with 31 per cent allowing posting of the version submitted to the journal, and 35 per cent not formally supporting archiving.¹⁴ Policies that allow author archiving are an especially effective way for scholarly societies to enable their members to improve the circulation of their work without endangering subscription revenues, and from the American Anthropological Association to the Society of In Vitro Biology there has been an embrace of this path to open access, just as there has been among commercial publishers from Elsevier to Wiley.¹⁵

Today, there are more than 2000 open access archives or repositories worldwide, with most of them maintained by university libraries, to which authors are posting versions of their work (OpenDOAR, 2012). While in principle this could lead to open access for at least half of the current literature, authors have not been exploiting this opportunity in large numbers. In response to relatively low levels of participation in self-archiving, a number of universities and research funding agencies have begun to mandate self-archiving in recognition of how it extends the value of the work.¹⁶ The leading instance of funder mandating is the US National Institutes of Health, whose Public Access Policy covers all of the research supported by the US$30 billion spent by this agency annually, which is expected to lead to 80,000 articles deposited in its open access archive PubMed Central (National Institutes of Health, 2008: 22; Turman, 2009). A recent study provides grounds for expecting the NIH Public Access Policy to have a positive impact on evidence-based medicine and healthcare more generally, given that between one-quarter and one-third of participants frequently accessed research literature, expressed an interest in having greater access and were aware of the policy and expected it to have an impact on their accessing research literature in the future (O’Keeffe et al., 2011).

The move to mandates has brought some of the more indirect economic matters of access to the fore. Publishers, including those who permit authors to archive, see mandated archiving as an incursion on the exclusive property rights they have secured from authors (in exchange for publication). They recognize the risk that this seemingly incidental open access introduces, and in some instances have curtailed an author’s archiving right, limiting it in many cases to the author’s final draft, after peer review (but prior to copy-editing, layout and proofreading), while also imposing an embargo period that prohibits archiving until 12–24 months have passed since original publication.¹⁷ Publisher organizations, representing the larger societies and commercial publishers, are also engaged in lobbying and public relations efforts that have resulted in the NIH adopting a 12-month embargo, and that more recently led to proposed legislation known as the Research Works Act, taken before the US Congress in 2011, which would have prohibited government agencies from mandating open access, in the way that the NIH has done (Library of Congress, 2011; Peak, 2008).¹⁸ The Research Works Act (now dormant) was symptomatic of continuing resistance to even limited delayed sharing. As the International Association of STM Publishers puts it, ‘publishers do not believe that self-archiving offers a sustainable alternative for scientific publishing’ (International Association of Scientific, Technical and Medical Publishers, 2008: 11).

Archiving is generally seen as a transitional and strategic state, involving the establishment of a parallel but lesser universe of scholarly communication that shadows the published record of the literature, while instantiating expectations around the value and viability of open access. Universities may use open source software to set up the archive within the scope of existing institutional technical infrastructures, but there are costs associated with operating them, while, at the same time, they do not reduce the need to subscribe to the published literature. The International Association of STM Publishers directs part of its attack on self-archiving by bringing up the expense of maintaining archives, citing a number of reports that place the costs of setting up and operating an archive to be between US$8500 and US$1.5 million, while calling for a cost-benefit analysis of operating archives (ibid.). The operating costs for arXiv.org are put at US$300,000 annually (Kenny, 2005).

There are also other sorts of costs associated with this form of archiving that offers open access to what is, in effect, degraded and delayed versions of the published record, setting up a coach class (archive) and a business class (journal subscription) access model that could end up taking hold rather than serving as a transitional stage to universal access to the published record. Archiving also has scholarly costs among those working in the humanities, where accurate citing of the final (published) text is essential to the very nature of analysis and critique, greatly reducing the value of self-archived versions, with similar concerns around precision of language making such scholars reluctant to post their drafts in any forum.

While author self-archiving remains the form of open access that applies across this otherwise stratified market, it was initially considered a marked point of success for open access that the large societies and commercial publishers adopted archiving policies, and it is this same segment which has now imposed greater restrictions on archiving in an effort to protect their intellectual property, while actively lobbying against archiving mandates among, for example, government agencies. Despite these costs and rising resistance among publishers, self-archiving certainly retains its champions who see it as ‘the faster, surer and already more heavily travelled’ road to open access, compared to the open access journal model, which we consider next, especially when that road is paved by institutional and government archiving mandates (Harnad et al., 2008: 39).

The archival vision of universal access to the literature, were it to be achieved by mandate and author compliance, is two-tiered. Researchers would have first-class immediate access to the journals to which their institution subscribed (as well as, of course, to open access journals), with archival access to everything else. Clearly, those working at less well-endowed institutions would benefit the most from this new state of access, even as it adds up to greater access for all. This prospect of mandated archiving, with its unknown implications for journal subscriptions, is certainly part of what is inspiring explorations by commercial publishers of open access options for their journals.

Open access journal publishing

If self-archiving extends across the three market segments with some evenness, the open access journal model has taken a different form for each of the segments. The Directory of Open Access Journals, as the best guide to open access titles, is approaching 10,000 in number (9937 in July 2013), having more than doubled since 2008, with examples from every discipline among the journals published independently, by scholarly societies and by commercial publishers. For independent journals, open access is proving the mode of choice as it places these titles from the moment on something of an equal footing to those of society and commercial publishers. Scholarly societies, on the other hand, are understandably torn between how open access might serve their members better by increasing access to their work while undermining one of the principal reasons for membership in the first place. They have tended to experiment with delayed open access, as a means of preserving the principle that membership should have its privileges. The commercial publishers (along with larger non-profit publishers in well-funded fields) are considering whether a simple transfer of rising subscription fees into rising article processing charges might prove the sustainable open access model for them. Let us consider each of these economies in a little more detail.

Open access independent journal publishing

With the move of independent journals to open access, I (JW) need to describe, by way of full disclosure, my own involvement in this segment of the publishing economy through my work with the Public Knowledge Project (PKP). Since 1998, PKP has sought to directly affect the economics of open access by providing open source software for journal management and publishing that can reduce start-up and publishing costs, while ensuring global online presence and indexing.¹⁹ Of the 11,000 journals that are currently using PKP’s Open Journal Systems software, we estimate that perhaps half are new, born-digital titles that have emerged over the last seven years, with almost all of them, old and new, providing open access to their content, although with a number doing so on a delayed basis, some time after initial publication, or while still selling subscriptions to their print editions (Edgar and Willinsky, 2010).²⁰

The economics of this form of open access depend on the sponsored and subsidized nature of scholarly communication, as the journals grow directly out of the resources and forms of work that are funded by the university, granting agencies and foundations. To a degree, the difference with independent journals is their apparent interest in simply sharing this work, rather than investing all that it takes to establish a business layer on top of this sponsored activity. Among the independent journals that I (JW) have worked with, Postcolonial Text was launched in 2003 without special funding or a source of revenue. It found the support it needed distributed across a fairly large community of committed academics who volunteered their services with the work. They have copy-edited, laid out and proofed articles, as well as managed a double-blind review process, in an effort to develop what might be termed the postcolonial quality of this interdisciplinary field, not only through open access publishing but also by assembling an international collective of editors, currently from Canada, India, Singapore, France, South Africa and Australia (Willinsky and Mendis, 2007). Open Medicine, to take a second example, was launched in 2007 by professional medical journal editors who sought to establish, by serving without remuneration, a new breed of medical journal that refuses the pharmaceutical and medical device advertising that otherwise sustains medical journal publishing, while offering open access to what they consider a public good (Willinsky et al., 2007).

A number of university libraries are setting a new course in scholarly communication by hosting this open source journal management software and putting up (independent) journals run by their faculty and students.²¹ In another instance of this new course, Simon Fraser University Library has taken the lead in the development of the software for PKP’s journal, conference and indexing systems. University presses are also becoming involved in new ways with independent journals, judging by the launch of Harvard University Press’s first open access journal, Journal of Legal Analysis, in 2009.

The independent open access journal is also proving a boon for scholarship in the Southern Hemisphere. Here, the International Network for the Availability of Scientific Publications (INASP), with whom PKP also works, has helped local editors set up regional online journal publishing systems in Vietnam, Nepal, the Philippines, Sri Lanka and Indonesia, with editors bringing their journals – some of them decades old – online within reach of national, regional and global indexes. The majority of these journals offer open access to their online editions, often going back to their first issue, enabling their articles to become part of a global knowledge exchange with a much broader reach due to open access.²² In Latin America, SciELO and Redalyc are two organizations which are providing journals with open access publishing platforms through which independent and society titles can also contribute to this global exchange.²³ The economics of open access in these cases builds not only on this new publishing medium, combined with the clerical-cost and energy reductions afforded by open source software systems, but on the formation of co-operative groups driven by common scholarly goals. To say they form an alternative to corporate market models would belie the contribution that Google Scholar, for example, plays by levelling the indexing playing field and increasing the global presence of this work. These are mixed economies (think, too, of the hardware involved), with the measure at issue not ideological fidelity but the means of affording increased access to knowledge.

Open access scholarly society journal publishing

With scholarly society publishers, increased access to research finds its champion through yet another research library venture, this time from Stanford University with the electronic publishing platform HighWire Press, which serves some 140 publishers – with a great many scholarly societies and non-profit publishers among them. The 1684 journals available through HighWire Press currently offer free access to more than two million articles (out of the nearly seven million articles available on the site), largely based on delayed open access, which typically involves releasing articles 12 months after publication. HighWire Press also helps its journals offer immediate open access to developing countries. Delayed and developing country open access offer scholarly societies a ready means of increasing readership, citation levels and the contribution of their members’ work, without putting the society’s subscription revenues at risk.²⁴

However, in selling subscriptions to libraries, society journals face increasing competition from the major commercial publishers whose journals, while more expensive, are often bundled in ways that take up an increasing proportion of the library’s acquisitions budget (Prosser, 2004). As noted, some societies find it easier to turn their titles over to a publisher rather than competing against it, which typically results in price increases, and profit sharing between publisher and society. The library community has responded by providing societies with alternative scenarios, including Crow’s business plan of forming publishing co-operatives among scholarly societies (2006). Such a co-operative could address the societies’ lack of market presence, undercapitalization, absence of staff and general risk aversion. For our part, we would like to see the co-operative idea taken a step further, with societies forming publishing co-operatives with research libraries as a more efficient and stable means of meeting the publishing costs of the society’s journal – compared to selling subscriptions. This would enable better integration of the journals into the library’s systems, with the co-operative’s journals then made open access as a benefit to both authors and the larger community (ibid.: 227–32).

Finally, a number of the larger societies working in well-funded research fields, such as the American Chemical Society, the American Society for Clinical Investigation, the American Physiological Society and the Royal Society of Chemistry, have begun to offer authors a chance to purchase open access for their individual articles within what is otherwise a subscription journal – a model that links these societies to the economic experiments currently underway among the major commercial publishers.

Commercial publishers’ open access

On 7 October 2008, the major publishers took a big step towards open access when BioMed Central, the largest for-profit open access publisher, announced that it had been bought by Springer, the second largest academic journal publisher, for a price estimated to be in the region of US$35 million (BioMed Central, 2008). Springer CEO Derk Haank summed up the move this way: ‘This acquisition reinforces the fact that we see open access publishing as a sustainable part of STM publishing and not an ideological crusade’ (ibid.). BioMed Central, which began in 2000 and refers to itself as the Open Access Publisher, offers close to 220 journals. To cover its publishing costs and generate a profit, it drew on the long-standing practice in the sciences of levying ‘page charges’ which were originally intended to supplement subscription fees.²⁵ In pioneering this approach to open access at the turn of the twenty-first century, BioMed Central initially charged authors US$500 to see an accepted article published, with institutional and national memberships providing another means of covering the charges. Meanwhile, what is now referred to as the standard article processing fee has since risen to US$1700 an article, with some variation among BioMed Central’s different journals (the highest fee being US$2290).²⁶

For its part, Springer had already begun experimenting with the article processing fee model with its Open Choice option, in which authors are able to purchase for US$3000 open access to their particular articles within an otherwise subscription-based journal. There are similar programmes underway within the other major publishers, Elsevier, Wiley, Taylor & Francis, and Sage.²⁷ While obviously not a commercial publisher, with over 240 journals Oxford University Press operates on a similar scale to the major commercial presses and, more importantly for our purposes here, has been far more forthcoming in providing take-up information about its open access option than the commercial publishers, even as it charges a comparable fee. Oxford Open is a US$3000 open access option that it offers for 90 of its titles. In 2006, 11 per cent of its authors in the life sciences took advantage of this option, with only 5 per cent in medicine and 2 per cent in the social sciences and humanities doing so (Richardson, 2006).²⁸ As a result of its experimentation with open access options, Oxford University Press has switched four of its science journals to completely open access titles, while adjusting subscription prices for the hybrid journals based on their open access revenues (Suber, 2007).²⁹ Whether this will form a pattern that the commercial publishers follow remains to be seen.

One interesting variation of the article processing fee might be termed the subdiscipline processing fee. The particle physics community has launched an experiment by the name of SCOAP³ (Sponsoring Consortium for Open Access Publishing in Particle Physics) which seeks to affect ‘a global conversion of the main corpus of [particle physics] journals to the open access model’ (SCOAP³ Working Party, 2007: 3). The group has calculated that with a US$15 million payment to publishers, open access to the entire field of particle physics could be purchased (based on a fee between US$1500 to US$3000 an article multiplied by the number of articles published in this field). The US$15 million turns out to be less than is being paid by libraries in total subscription fees for the six relevant journals in which over 80 per cent of this literature appears (Mele et al., 2006). Libraries are being asked at this point to commit to shifting their subscription fees to this purchase of open access, with the actual amounts committed based on their country’s share of articles produced. To date, SCOAP³ has managed to secure pledges for roughly half of the amount needed from libraries and other agencies in 18 countries, while the publishers have expressed an interest in tendering bids on this whole-field open access purchase. With the price of physics journals at the high end among the disciplines (Kean, 2007), this sort of direct negotiation seems a positive step, especially in conjunction with an innovative form of co-operative made up of the libraries and particle physics community that seeks to establish, for the benefit of the wider world, open access to its literature.

It also needs to be said that an article processing fee repositions the transaction between publisher and author. Rather than authors turning over their copyright to publishers in exchange for publication, authors are contracting a service from publishers that ensures the widest possible distribution of their work under the journal’s imprimatur. This advances the researchers’ standing within the academy’s reputation economy, in which tenure, promotion and relocation depend on getting one’s name out there. With the subscription model, on the other hand, publishers balance journal prices (and price increases) against subscription sales (and loss), limiting the reach and impact of their authors.

Yet this contracting of publishing services only sustains the stratified costs among the three segments of the market, which will continue to add up to a bill that the academic community will struggle to meet. With commercial publishers (as the most expensive segment of this market by far) increasing both their market share and their interest in open access revenue models, the academic community is facing an odd future of open access to the research literature potentially costing more than the current mix of subscription-based journals.

How can a US$3000 article processing fee be justified, while thousands of open access journals are publishing peer-reviewed articles at no charge to authors or readers? The sustainability that appears to be at issue with open access seems to be about sustaining current revenue and cost differences that otherwise appear unwarranted. To be fair, the International Association of STM Publishing has provided a series of explanations for the value its member publishers add to the articles they publish. The association points to ‘the collective investment of hundreds of millions of euros in electronic developments’, as well as highlighting how ‘where editorial administration takes place within the university, it is usual for the costs to be charged back to the publisher’. It reminds readers that ‘profits are a major source of reinvestment and innovation’ while ‘society publishers frequently use surpluses from journals to support other activities such as conferences, travel and research grants or public education’. Then there is the ‘3–7 year investments in the development of new journals around which emerging scientific communities seek to coalesce’ as well as the additional review management costs for those journals that have a very high rejection rate, when only accepted articles will pay the article processing fee (International Association of Scientific, Technical and Medical Publishers, 2008: 6–8).

All that is missing from these explanations is any acknowledgement that these same publishing services are being provided at far less cost by the academic community itself. Those who edit independent journals, as well as the journals of smaller scholarly societies, provide ‘registering, certifying, formalizing, improving, disseminating, preserving, and using scientific information’ (ibid.: 3). They launch new titles and co-operatively develop publishing systems, and are no less committed to ‘making long-term investment in publications around which emerging and established scientific communities coalesce and evolve’ (ibid.). They are finding ways of offering open access through co-operative efforts and shared capacities that represent innovative and efficient approaches to scholarly publishing, which can, through integrated data sharing, for example, add to the quality of the resulting knowledge.³⁰

Conclusion

What is missing from the current economics of open access is a more exact accounting for pricing differences and why they should be sustained. Can we begin now to justify and rationalize historical differences in editors’ remuneration and support, resources devoted to acquiring and promoting titles, operating costs and corporate structures, peer-review management, in profit and surplus levels? Can we sort out the requisite services, the balancing of private and public interests, commercial and co-operative ventures, the basis of ownership and distribution, to ensure that monopolies granted to publishers (by authors transferring copyright to them) can be said ‘to promote the progress of science and useful arts’ – to quote the US Constitution – to the greatest extent now possible?

The need for a formal reckoning, at this moment of digital transformation, could not be clearer, given the tussle between publishers wishing to sustain an economy that the academic community finds unsustainable, on the one hand, and the commonly recognized potential for far greater access, on the other. There is a sense that the academic community has accumulated the resources and capacities, in co-operation with commercial ventures, to greatly reduce the restrictions on who has access to what knowledge, and to approach universal access to this body of research and scholarship. Given what is at stake, one has to wonder if the economics of scholarly publishing are not worthy of a US congressional hearing, or some other means of working this out.

While it seems unreasonable to expect major commercial publishers to abandon their current advantage, having worked so hard and so smartly, perhaps recent economic hardship will alter such corporate entitlement rights. A downturn is bound to hit library budgets no less than the rest of the economy, just as the extent of this recession may well lead governments to alter the regulation and protection of such areas of public trust. As things stand, the academic community needs to insist that the major publishing organizations are not telling the whole or the only story when it comes to scholarly publishing. By the same token, a number of us will continue working on providing the means for, and research on, an economics of open access that: (a) is facilitated by shared and open source resources and co-operative ventures involving societies, research libraries, university presses, and others; (b) develops capacities for scholarly publishing in developing countries; (c) supports path-breaking scholars seeking to set new fields in motion; and (d) offers scholarly societies a viable alternative for continuing to publish their own journals, particularly those working in low-budget and under-funded research areas where substantial article processing fees are beyond the pale.

The fundamental question that scholarly publishing faces going forward is not whether open access is sustainable. Rather, the sustainability question that scholarly publishing faces now and into the future is far more about the economic stratification that emerged in the years leading up to online publishing and open access, and whether it should continue unexamined and unchallenged into the digital era. A change of publishing mediums on this historical scale typically alters the communication economy, changing production modes, audience profile and content form. Thus, it seems reasonable to expect a greater rationalization of the billions invested in scholarly publishing, a rationalization that would best begin, this chapter suggests, by challenging the need for current levels of economic stratification while seeking to increase the openness with which this public good is cultivated, circulated and built upon.

Acknowledgement

This chapter is an updated and revised version of an article by Willinsky (2009).